DeepSeek search and fetch – $0.75 per 1000 successful requests

416 pointsposted 10 hours ago
by fagnerbrack
(api-docs.deepseek.com)

180 Comments

nikvdp

8 hours ago

What are people actually using these hosted search providers for? For personal use I've found using searxng-ai-kit [1] (disclaimer: I'm the author) or similar tools to be far more cost effective (free!). oh-my-pi uses a similar method and will scrape search results for you out of the box, no api key or search provider needed.

[1]: https://github.com/nikvdp/searxng-ai-kit

hobofan

7 hours ago

Any organization that wants reliable latency, data processing guarantees (it's a big leak vector for company secrets), etc., will use such a search provider over the solutions you outlined.

Especially for chat-like use-cases (in contrast to coding), latency is a lot more critical, especially if you kick off a small single query question. On antother axis, scrape-reliabilty is important for a deep research job that will fire ~10+ web searches and ~100+ web fetches.

e.g. in our chat platform[0], we offer first party MCP servers for web search[1] and web access[2] which work decently well, but some of our customers that prioritize quality over cost, they go with dedicated providers like LinkUp[3]

[0]: https://github.com/EratoLab/erato

[1]: https://github.com/EratoLab/web-search-mcp

[2]: https://github.com/EratoLab/web-access-mcp

[3]: https://www.linkup.so/

ryan_glass

4 hours ago

Can you explain how this would help protect a company's secrets? I would think that if you want to keep your IP secret surely using something like Deepseek's API directly is not ideal and hosting something locally is best practice.

hobofan

2 hours ago

I personally also wouldn't trust Deepseek with companies secrets. That's why I linked LinkUp, which I know has passed through a quite strict compliance process at a few of our customers.

Obviously, hosting a search index locally would be best practice, for the search part, but usually comes with quite a significant hardware cost and operational complexity, if you want to have an up-to-date index. I'm not aware of good self-hostable options in the space but would be happy to hear about them if anyone here has experience with that! AFAIK LinkUp and some other provides do offer private cloud/on-premise deployments, but I think mostly for their biggest customers.

For the web fetch/access part of the equation, that is inherently live and not containable for a local deployment (unless you would host a whole internet archive/Wayback Machine, which is even more of an undertaking). There, the main options you have to prevent leaks, are a domain allowlist (if feasible), doing query parameter stripping (error prone, and still no 100% guarantee), or configuring the client (e.g. our tool) in a way that users have to explicitly allow those tool calls, and are informed about the posed risks. We are also working on a system, where certain tools are fully deactivated once e.g. an office document that's marked as confidential enters the context.

For the highest requirements on company secrets, an air-gapped system that doesn't have web search/access capabilities is obviously the ultimate solution.

visarga

5 minutes ago

Why can't we simply save search stubs, short documents a LLM can retrieve, use them to route to various websites, resources, feeds, forums, blogs, other specialized search engines, APIs? If you host a 30GB model, then you can use 30 more GB for a compressed search stub index. The entry points for any domain are more stable than the pages. You would use an agent to navigate the resources as needed, fetching fresh data.

jeromechoo

2 hours ago

Diffbot might be the only company that offers a self-hostable fully airgapped web index with access to content cache (ala archive.org). I work here, happy to share more info about how it works!

jgalt212

an hour ago

I would use this to discover leaks, not to protect data.

Havoc

6 hours ago

I have a searxng instances but stopped using it after doing a comparison vs actual search provider APIs.

searxng gets rate limited fast and the results are inconsistent depending on what providers it can reach at any given time / isn't captcha'd out.

Meanwhile there are multiple providers with generous free tiers so no point in feeding LLMs inferior data

Also did a couple of side by side tests asking LLMs directly which feed results it prefer for some sample technical searches. Exa and tavily #1, brave #2, searx last

xienze

4 hours ago

> searxng gets rate limited fast and the results are inconsistent depending on what providers it can reach at any given time / isn't captcha'd out.

Exactly. I really wonder what the people going on about using SearXNG are actually doing, because in practice I've seen just what you see. Even from a residential IP, you might get a handful of Google searches before getting backed off. If you ever have multiple searches going on at once, even Bing and Brave are going to back you off pretty quick. My home IP is apparently permanently banned from using Brave Search through SearXNG.

And when you aren't getting results from Google/Bing/Brave, it gets bad really quick for your agent. As much as I wanted to go fully self-hosted for agent search, it just doesn't work well.

mosselman

6 hours ago

I have built my own search engine, which is really great btw. I self-host it.

I’ve combined multiple paid search providers and some that don’t have APIs. These API-less providers I’ve written wrappers for that scrape the normal html pages and return objects from the results.

These wrapped search providers very often block the engine from getting results. Either through rate limits or by blocking my server’s IP.

I am the only user and still get blocked. Let alone if multiple people would hit let’s say DuckDuckGo from the same office IP all day or in a long research session that requires many look ups.

An obvious solution is using a proxy with automatic rotation or some sort of captcha breaking system, which is fine for an agent, but in a search engine scenario might up costing too much in latency.

lxgr

3 hours ago

> I am the only user and still get blocked.

I’m not surprised. Google these days regularly blocks me, a human user with a Google account with decades of tenure.

apwheele

4 hours ago

I am currently using Perplexity for https://veruscite.com/. Verify citations actually exist -- so get in a list of 100, half I can check with crossref, but the other half I need to do a web search and fetch to make sure they exist.

So I need higher rate limits to not make the tool come to a crawl. I probably will not consider DeepSeek API here (I have tried DeepSeek hosted at US providers, like Perplexity). The other search/fetch tools I try do not have great web coverage of Asian journals, but the fact DeepSeek API is not zero data retention I cannot get over.

thombles

6 hours ago

For OpenWebUI and also OMP with local models. Searxng is, uh, playing fast and loose[1]. More engines are implementing anubis or other challenges, which searxng attempts to implement/bypass, but they are also making these services inactive by default to limit load[2].

Once this cat-and-mouse game goes on a bit longer I expect the quality of free searxng results to crater. My advice to anyone who has any budget at all is to skip over this whole issue and spend a few bucks on a proper search API.

[1] https://github.com/searx/searx/issues/1237 [2] https://github.com/searxng/searxng/pull/6696/changes#diff-e9...

oefrha

6 hours ago

IMO unless you refuse to play the LLM game altogether (then you wouldn’t be having this discussion in the first place), at the end of the day you’re using tech built on top of disrespecting every IP claim and TOS out there on a fundamental level, so no one’s clean and it’s only a matter of how much you can deceive yourself into believing you’re clean.

On a practical level though, using searxng and co. and playing the game of cat and mouse with providers that really don’t want you to use it is just not worth it unless $0.75 per 1000 searches is really expensive to you or you really need the anonymity.

thombles

5 hours ago

I take your point, but there remains a difference between a multi-GB file sitting on my hard drive and my use of a large complicated database fielding my search queries. I don’t believe that anybody is obligated to run a search engine for free for my benefit and it’s logical to pay a fraction of a cent if that makes it work.

dofi4ka

7 hours ago

Hosted search API from your LLM's provider is much easier to use, faster and less error prone. Also I believe that LLM may work better with its provider's search api, as it was definitely trained on it. I mean, it is just more convenient to use your provider's search.

And if we talking about LLMs, there is no particular point for search to be absolutely free if you still pay for the LLM. It would have more sense if search AND llm would be completely free.

lxgr

3 hours ago

Self-hosting LLMs is much more feasible than self-hosting search, so that seems like a relevant use case.

ATMLOTTOBEER

4 hours ago

The only reason this exists today is because normal search providers haven’t been completely disintermediated.

As soon as 100% of traffic to Google searches is not human eyes, they can’t fund it by ad rev, and you will need a service like this

Faaak

7 hours ago

How do you deal with fcking cloudflare blocking access to the websites ?

mjmvisser

3 hours ago

These services only exist because Google refuses to provide a paid search API.

prism56

3 hours ago

Thought they'd be allover it. Dropping in cheeky ad results.

techscruggs

2 hours ago

Everyone would just programmatically filter out the ads.

ijidak

2 hours ago

Have never understood this.

A paid search API seems like it would be easy money for Google. And maybe some lock-in for all sorts of start-ups and enterprises.

There are so many apps that would benefit from search.

I get that it means they can't serve ads.

But you've got to be willing to disrupt yourself or you'll lose eventually anyway.

Better to figure out how to make money from ads and an API, than to cower in fear.

That same fear may be why they failed to be first movers in LLMs.

Have to be willing to disrupt yourself.

At the same time, I've never built a billion dollar company. So I'm absolutely armchair quarterbacking here.

thombles

9 hours ago

For comparison that’s $5 with Brave Search API and $12 with Kagi API

wmchen

8 hours ago

Ceramic.ai is $0.25/1000 searches. I've been using it as an MCP in OpenCode and it seems adequate. I don't see many people talk about it, and I have not benchmarked it against Exa/Tavilly.

breadislove

2 hours ago

ceramic is very very bad, the worst search provider of all. it feels like it is doing only lexical matching and the index is very outdated

vlmutolo

2 hours ago

They explicitly are only doing lexical matching. I actually don’t mind this if it lowers the cost; LLMs can turn lexical search into a very powerful tool by iterating on the queries. Just look at the success of coding agents using grep and similar code search tools.

But it does seem like their index isn’t great.

loehnsberg

7 hours ago

Kagi MCP works really well imo also within DeepSeek Harness. Just looking at cost neglects the possible difference in quality. How good is that DeepSeek web search/fetch?

ece

9 hours ago

$4 with Exa, $1.50 with SERP API

ttul

an hour ago

SerpAPI is comically good at evading Google's scraping protections. If what you want to know is, precisely, what someone searching Google in Pakistan would see, this is the API for you.

stncls

7 hours ago

Mojeek is £ 2 per thousand. They have their own index (not a SERP), but subjectively pretty close to Google (I prefer it to Bing/DDG).

hobofan

8 hours ago

$1 for LinkUp fetch, $5 for LinkUp search.

satirr

7 hours ago

is the cost of search and scrape provider is impact the quality too? for the search is yes, because there it shows the quality and depth of the search. but for scrape, wht the cons?

Doohickey-d

10 hours ago

How well does this (and similar services) deal with anti-bot protection? What happens if I attempt to retrieve a page protected by the cloudflare checkbox or by Anubis? An error, no useful content, or the actual content I want?

tyre

9 hours ago

> Search API finds web pages for a query and returns structured results. It is powered by DeepSeek's in-house search engine.

Looks like they've built (or are building) their own index and return cached results. What you do with that information is up to you. If you want to visit the live page, you'd need to find a way around bot detection.

One reason (not the only one!) people use Anubis is to prevent the endless flood of requests from AI companies. If Deepseek scrapes once every few hours, that helps the website owner with the traffic problem while allowing access.

If they don't want to allow access, my guess is that deepseek's search crawler respects robots.txt but I don't know.

spiderfarmer

7 hours ago

I have always blamed Chinese AI operators for the hundreds of thousands of residential proxy requests I receive each day, but there's no way to be sure.

They absolutely dwarf real traffic if left unchecked (I blocked all countries outside my target area).

nativeit

3 hours ago

If there’s no way to be sure, then why blame Chinese operators? Because American companies have given you so many reasons to believe they’re acting in good faith?

I don’t particularly trust anyone at this point, and I assume AI traffic on my web servers are all malicious actors, and never bothered to consider what country the malicious bots came from. Mostly because I couldn’t possibly care less, but it kept happening even with geofencing and rate limits, so some American companies who own domestic IP blocks are facilitating them.

MintsJohn

3 hours ago

Identified AI bots that finally made me cave and put anubis (with very liberal limits) up were anthropic/openai/perplexity according to the headers they sent. They were also over 99% of traffic despite proper robots.txt (site only has a few hundredth visits per day), which they ignore. I don't mind bots or crawlers, but these bots got the same page/query string every time they decided it was time for a crawl, which was about daily, cache control, last modified headers they also never seemed to be interested in. They were very very far from well behaved, unlike search engine crawls, and impossible to tame.

vincnetas

6 hours ago

Yeah, but how these residential proxies get in the residential areas? I would guess popular VPN providers are mostly behind this. They sell you VPN and sell bots residential proxy access on the other end.

esseph

4 hours ago

Third party. A third party provides endpoints (hacked routers, phones, laptops, desktops, IoT devices, etc.)

weird-eye-issue

10 hours ago

There is no single answer to this, it entirely depends on the exact service and the exact page

Most US services like Claude and ChatGPT send a specific user agent that makes it easy to block them and specifically does not try to get around any sort of bot protections

Meanwhile other services could use a variety of tactics such as simply setting a browser user agent and/or using random data center IPs or residential IPs, etc etc and then the blocking would just come down to whatever the individual page has configured which can actually be pretty hit or miss

bulder

8 hours ago

From what I've seen, when operating in an environment with bash tools, Claude and ChatGPT will react to a site blocking their normal, attributed fetch tool by calling curl or writing a python script with a different user agent to try and bypass the blocks, rather than respecting the site owners' wishes.

It'd be one thing if they required the user to consciously say "yes, I want you to bypass the block" (still unethical), but the fact that it's just default behavior really says something about the service design alignment.

gvfsa

7 hours ago

Claude and ChatGPT are user agents. For obvious reasons, they serve the user, not the websites' owners. (Also, if you don't want people to be able to access some information, then don't put it online.)

bulder

4 hours ago

My point is that they're not serving the user in that case, since they're taking actions that bypass site terms without consulting the user. And for example, if the site forbids scraping, that attempted block bypass can get the user banned without ever seeing any warnings.

matkoniecz

7 hours ago

> if you don't want people to be able to access some information

Claude and ChatGPT are not people

gvfsa

6 hours ago

Chrome and Firefox aren’t people either.

pseudosaid

6 hours ago

they are tools. who do you think is using them? dogs?

your comment is another classic hacker news comment. missing the meaning to validate themselves. Lower than my own effort, but wow does it come in so many forms on this site. as if mine is any better

esseph

4 hours ago

> they are tools. who do you think is using them? dogs?

My agents are using headless chrome :P

weird-eye-issue

4 hours ago

Yeah I should have clarified I was primarily referring to the standard web interfaces, however since pretty much all of these tools can execute code these days it is true that technically they can and often will work around blocks. The degree to which this is intentional isn't super clear - if they locked it down too much then it might start refusing to write code that does any sort of web requests which would obviously be an issue for just standard development tasks so it's not super straightforward. But it is true that the built-in tools that they have access to to fetch pages have a hard coded user agent that identifies them

bulder

4 hours ago

Fair enough yeah, I think the thing that is annoying is that the service design is explicitly disincentivizing keeping an eye on the tasks, with all of the "set it and forget it" features. You really have to put in work to make sure that the model takes breaks to consult you instead of making assumptions (which in OpenAI's internal testing seem to be 'the user wants me to break into a site').

skeledrew

4 hours ago

If you hire someone to handle some task for you, do you want them to keep running back to you every time they run into a blocker that they could potentially resolve on their own?

nativeit

3 hours ago

That depends if they routinely and cheerfully behave in objectively insane ways.

skeledrew

3 hours ago

> objectively insane

Contradiction: one person's insanity is another person's genius. And yet another's "Meh, this is the norm".

bulder

4 hours ago

I'd prefer my little subcontractor to respect the commons on which I depend.

gvfsa

4 hours ago

In what way does loading a website using the Claude Code harness not respect the commons, as opposed to using Firefox? If anything, using Claude Code is better, since it doesn't load the unnecessary static files (javascript, css, etc), saving the website CPU, I/O, and bandwidth.

skeledrew

4 hours ago

Are you still paying them for the time they spent when the task you gave them remains incomplete, or you end up spending the time/effort you hoped to avoid spending by hiring them in the first place?

minraws

8 hours ago

This is not true, Claude and ChatGPT have sent traffic to sites I maintain without any user-agent, it might be they have stopped doing this.

Google has done various versions of this as well, Grok too.

It could be these bots were in some custom harness or something, but they literally sent requests in very weird ways, that humans won't be doing.

I have ended up putting my sites behind vpns or anubis and similar. This is beyond painful.

ReptileMan

8 hours ago

>Most US services like Claude and ChatGPT send a specific user agent that makes it easy to block them and specifically does not try to get around any sort of bot protections

The last part is extremely dubious with some recent reveals.

jack_pp

8 hours ago

recent reveals show that OAI and Anthropic give users the ability to go around bot protections or they internally use it?

amluto

8 hours ago

It’s not exactly recent or a reveal, but I’ve caught at least several OAI models writing Python scripts to access web pages.

skeledrew

4 hours ago

If you the user ask it to access a page (or do any task that probably needs that access), then it'll try to do so the best it can. Anything less is user-unfriendly behavior.

amluto

2 hours ago

I disagree. If I turn off web search in the config, then the tool should not search the web, thank you very much. Maybe if I very explicitly ask it to, then it should try.

The whole succeed-at-all-costs idea is a big part of why there have been so many massive security screwups lately.

skeledrew

2 hours ago

Well if you disable web search, that's just making that built-in tool unavailable to the agent (when it's enabled, the agent gets usage info and guidance about it in its context). Given a task in which online resources may be useful, the only thing the agent knows is that it could search the web, and that there are a variety of means (curl, wget, urllib in python, etc) which may be at its disposal. Without you telling it specifically not to search the web (or that guidance coming from elsewhere) or blocking internet access, there's no reason it wouldn't use that avenue in fulfilling its task.

amluto

an hour ago

I am well aware of why this phenomenon happens. That does not mean that I think that OpenAI’s (and Anthropic, etc’s) training or harness behavior is appropriate.

Would it really have been so hard to encode an indication of whether use of the public Internet is permissible in the prompt and to train the models to not attempt to exceed authorized access?

skeledrew

38 minutes ago

I'd say for the greater majority of users, when they tell their agent to do some task, particularly any that involves research (which isn't specifically about something on the local machine), using the internet is implicitly expected. Imagine having to go into my computer settings to flip an "allow internet usage" each time I want to use a browser. That's extremely user-unfriendly, and there's 0 reason to encode a default internet block when users expect their agent to handle their given tasks with minimal friction (I still recall the frustration of babysitting Claude permissions just a few months ago; auto-mode was the best feature ever for those of us who didn't want to go YOLO).

ReptileMan

6 hours ago

I have noticed that agents become extremely creative to fulfill any non refused requests. And then instructions become more of a suggestions.

m00dy

9 hours ago

I've spent a lot of time building datamint.xyz, and dealing with user agents, datacenter IPs, and VPNs gets pretty tricky. What works right now is a fork of the latest Firefox with anti-fingerprinting patches, paired with a high-quality residential proxy network with clean IPs. Even the basic setup needs 11+ code-level patches, and there are still a few edge cases, but this setup should cover 99% of cases.

thisismyanon2

9 hours ago

Thank you for escalating the arms race that's currently destroying the internet. I can't look at the Apple user forums, my health insurance company's portal, pay my county taxes, browse Digikey and Mouser, view Github or whatever forge, etc. etc. without having to jump through Cloudflare or Anubis garbage designed to block your abusive behavior.

I'm just overflowing with joy that you're comfortable wasting my time and money.

lukewarm707

6 hours ago

I use the automated captcha solvers, stealth browsers, proxy IPs and so on. In regard to ethics, I think this is the right thing to do. The minimum I can do is just to show this is not a matter of excuses or mental gymnastics to trick myself into thinking it's fine.

Even if you said, a lot of the public commons is destroyed, I would NOT use bots and cause more destruction because 'everyone else does it' or 'it's a lost cause'. I say it is the right thing to do because the balance is such that using the internet is actively harmful to the user. That is, using the internet with bots is neutral to the user, and using it without bots, you are being actively harmed and exploited.

If you are using search, this is not a library. You are using your time to train targeted advertising that exploits you; you are selling your search history to the highest bidder. If you are using ChatGPT you are just working as a teacher for free.

On services, you have paid for the product, but the seller is scamming you by not giving you the product. The seller is using dark pattern UI to trick you.

If you are on social media, don't think of this as a public park. You are being treated like a lab animal, your behavior is being studied for the benefit of a corporation.

So the point about the bots is that I think people should be empowered to protect themselves from the internet, which is otherwise actively harmful to their wellbeing.

stavros

8 hours ago

Even if it's not the GP, if there's profit to be made, somebody will do it. I was reminded of the power of economic motives the other day, when it turned out someone bought an entire trustworthy reseller just so they could compromise the Ledger supply chain by including a tiny eSIM module that OCRed the screen and transmitted the seed.

Incentives are crazy powerful.

matkoniecz

7 hours ago

> if there's profit to be made, somebody will do it

Not always! With sufficiently powerful barriers it can be blocked at least in part.

Shaming awful people is one of available tools.

Or at least not praising them for making small profit at huge costs to others.

stavros

6 hours ago

Can't argue with doing what we can.

stev1983

8 hours ago

Here’s what is says on their site: “Courtesy backoff Adaptive backoff keyed to target response codes. When a target returns 429 or 503, we back off immediately, so our fleet never applies denial-of-service pressure to a site.” By the time you get the error you have already crashed the site. Geez.

lukewarm707

6 hours ago

"Usage records — which endpoint was called, when, the credit cost, the status, and the target URL."

I would use it, but then the privacy policy says that you would store my search history. I simply want to use the internet in peace without being constantly recorded.

m00dy

6 hours ago

hey,

I didn't get your point, how would you do design our usage records then ?

lukewarm707

5 hours ago

Ideally it would be private.

Store: Customer ID, billing details, services/credit used.

Try not to store: The user's search queries, the user's prompt/response content, the user's web history (URLs, etc).

To avoid selling the llm prompt content you would need your own inference or a contract.

Sadly most just hoard the data and some will offer a model where they charge for privacy. They are all quite scammy.

Firecrawl: logs. Parallel: logs. Bright data: logs. Kagi: logs for abuse. Bocha, querit etc: logs logs logs. Cloudflare: doesn't say what they do. Exa: charge for no retention. Linkup: no retention on request. Brave: no retention on request, but actually approved it (win for brave).

latexr

9 hours ago

I have to agree with the (sarcastic) sibling comment. You’re essentially making everyone’s life worse in the name of personal profit.

Please stop. The last thing we need in the world is more greed and selfishness.

lejalv

8 hours ago

Since when have moral appeals worked against the central feature of capitalism?

Only strong regulation can help.

matkoniecz

7 hours ago

Shaming awful people is one of available tools.

Or at least not praising them for making small profit at huge costs to others.

Also, it is useful as coordination method, enabling passing laws outlawing some behaviors.

(and I am not able to legislate-via-http while posting such comment is easy)

luckylion

5 hours ago

Anti-fingerprinting is mostly there to be able to use the same browser across many requests, I assume?

I found that a slightly-modified firefox for more stealthy remote-control works fine and gets a pass on all automatic captchas I tested. but I don't do it at scale, I just wanted something to automate a particularly stupid website that I have to use.

Would love to see if my browser is detected -- any recommendations for tests to run?

czk

2 hours ago

usually a combination of something like camoufox and a residential ip address

m00dy

10 hours ago

you can bypass those easily except Google Captcha v3, you need to have your own proxy swarm to make it at scale.

frenchtoast8

9 hours ago

For comparison, Kagi charges $12 per 1000 searches (only talking about their API pricing, like what you’d use with an LLM; their $10/mo plan has unlimited searches but it’s a TOS violation to use those credentials with an LLM).

Havoc

9 hours ago

>it’s a TOS violation

Competitor uruky is more relaxed about api and is cheaper. Creator is active on hn

https://news.ycombinator.com/user?id=BrunoBernardino

wmchen

7 hours ago

Thanks for sharing, wasn't aware of this provider. If anyone is wondering about rate limits for the subscription:

>We calculate search rates based on the timestamps of the last searches of a user (just the timestamps, not the queries or anything else!). We keep those timestamps for 12 hours, then discard them. We currently limit searches to 5 per 5 seconds, 15 per minute, and 120 per hour.

Source: https://uruky.com/faq

API: https://uruky.com/docs/api

In other words, good enough for casual use if you were already going to subscribe. Assuming the rate limits are okay, Uruky’s €5/month becomes cheaper at these search volumes:

1. Tavily Basic: 1,704 searches/month

2. Brave Search: 2,126 searches/month

3. Exa Fast / Auto: 2,233 searches/month

4. Exa Instant: 3,907 searches/month

5. DeepSeek Search: 7,503 searches/month

6. Ceramic: 22,507 searches/month

Havoc

6 hours ago

I ended up building an MCP that pulls results from multiple places and dedups it.

The free tiers on most are generous enough for normal agent MCP usage

Also...pretty sure decodo just launched a search thingie

BrunoBernardino

6 hours ago

Wow! Thank you so much for sharing! Happy to answer any questions about it.

PhilippGille

8 hours ago

Kagi has an official MCP server [1], where you set the API key. MCP use is always by an LLM, so this seems to be at odds with their TOS.

[1] https://github.com/kagisearch/kagimcp

wmchen

7 hours ago

I believe the commenter was referring specifically to the paid subscription for Kagi, not the per-query usage-based billing API.

suncemoje

9 hours ago

About a year ago I built my own search and fetch MCP tools for agents based on third-party scraping providers. It is about 15-20x more cost-effective than what OpenAI and Anthropic is offering with their integrated web search and fetch tools. Very impressive that DeepSeek has such a competitive offering today.

aurareturn

8 hours ago

Source for those scraping providers?

suncemoje

7 hours ago

ScraperAPI and ScrapingDog. The latter also offers Google Search at a fraction of the official Google API price. Wrap their APIs into an agent/MCP tool et voilà.

himata4113

an hour ago

I wonder if providers do this so they can 'scrape' for 'free' because the customers pay for it? They all seem to have an abstraction around search rather than suggesting the use of native search apis.

dsco

10 hours ago

DeepSeek Flash really came at a bad time before the Anthropic IPO. I use this thing daily and it’s neck and neck with SOTA models. The search is one of the things setting it apart.

rjh29

9 hours ago

It's absolutely not SOTA for vibe coding a whole app, but for guided work it is great.

platinumrad

9 hours ago

Is anyone actually oneshotting a whole app or even vibecoding one in multiple steps without it being "guided work"? I really wish announcements (like DeepSeek's recent one) would stop focusing on the ability of the model to spit out Minecraft clones on demand.

antoniojtorres

9 hours ago

Some people are trying to one shot entire companies

bahmboo

8 hours ago

And failing.

IshKebab

8 hours ago

Someone at work vibe coded a replacement for a entire third party SaaS that we were using in a week. Their version is better.

It wasn't one-shotted of course but I don't think there's much difference in the outcome between one-shot and like 100-shot over a week. That company is still doomed.

Think of something like Trello. Who is going to pay for that when you can easily vibe code something superior (yes you really can now).

rjh29

4 hours ago

Open source companies somehow make money off enterprise contracts. People still don't want to be on-call if something breaks, or have to manage servers and backups. But making money off SLAs to implement new features will probably be a thing of the past.

IshKebab

3 hours ago

Indeed, but only a tiny handful of companies have been successful with that business model, and they mostly make margin on things you'd have to pay for anyway like cloud services. No company is going to pay e.g. $15/user/month for Jira when they can get the same thing for free - better even. And yes it isn't an exaggeration to say you could vibe code something better than Jira. (Especially because that's a low bar!).

I'm kind of surprised nobody has done that yet actually. I guess Opus 5.5 has only been out for a very short time...

muhfuhkuh

an hour ago

The Red Hat or Canonical of vibe-coded/decomp/cloned software is just around the corner.

Slop curators and tech support shoulders to cry on. Could be lucrative.

MrScruff

9 hours ago

OP didn’t mention one-shotting.

There are plenty of use cases for the ability to quickly through together and refine an application without reviewing the code though.

gcanyon

9 hours ago

I have one-shotted the two most recent relatively complex apps I've tried:

1. a puzzle game with complex interplay, pieces to place, design vs. play mode, storing multiple solutions with the puzzle, and more. I've since iterated something like 200 times as I worked out the design etc., but the first iteration was 99% correct and what I asked for

2. a learn-to-read application with built-in stories (that the LLM wrote) the ability to import epub, text, and HTML, that uses the built-in speech libraries as a default, but also supports an open-source TTS library that I gave in the first prompt, tracks progress and problem workds, has multiple games to reinforce learning, and has a parental mode for tracking progress and configuring the tool overall.

Frankly I was stunned that Claude one-shotted that second app, there was a lot in the prompt, and it did it all. I fully expected to have to iterate on the first draft multiple times. But it worked out of the box.

IshKebab

8 hours ago

With Opus 5.5 you can vibe code a whole app without reading the code at all. Obviously you still have to tell it what to make.

With earlier models that doesn't really work. You really have to read the code and watch it like a hawk or it will make a big mess.

To put it another way, with lesser models you have to guide them at a code level. With Opus 5.5 you only have to guide it at an architectural / feature level.

microtonal

8 hours ago

People said the same for 5.0, 4.8, 4.7, and...

I am not disputing that the models are getting better for vibe coding. But I think people have very different standards of what a 'big mess' or 'maintainable software' is.

himata4113

an hour ago

The only meaningful improvements were opus 4.5 and now opus 5.5, people claiming otherwise are delusional. The gap between opus 4.5 and the next big thing was massive, now the same is true for opus 5.5 as much as I hate to admit it anthropic is running the show right now.

IshKebab

3 hours ago

They did but IMO they had very low standards. I have very high standards and Opus 5.5 pretty much meets them (at least for Rust).

"People said it before and it wasn't true then" isn't really an argument about it being true now. There are plenty of things that happened where the zealots jumped the gun with the hype. Bitcoin for example. "People said bitcoin was going to succeed in 2005, 2006, 2007...". Or even the web. It was mega hyped in the dot com boom and then crashed and of course loaaads of stupid people thought that was the end of it.

throwaw12

9 hours ago

which model is SOTA for vibe coding a whole app? I always use models+agents in guided setup

the moment there is a model which can code whole app, we can fairly say software engineering is done.

IshKebab

8 hours ago

Opus 5.5 + Claude Code (or probably any agent harness; I don't think they make that much difference). I wouldn't say software engineering is done quite yet but yeah it's definitely never going to be the same.

redox99

10 hours ago

It's really not neck and neck with Opus 5.5.

lejalv

8 hours ago

Cost-adjusted,it might be ahead.

It's key to name the boundary conditions. Unlimited cost is a very HN assumption.

TomGarden

8 hours ago

Max Subscription seems to be the HN assumption most of the time

Rodmine

5 hours ago

True, it isn't neck-and-neck. It's neck and balls. Deepseek's balls and Opus 5.5's neck. Whatever you can build with Opus 5.5, I can build in the same time, 1/500th the price with Deepseek Flash 4.1.

eru

9 hours ago

Really depends on your use case.

You are right that there's plenty of use cases where Opus 5.5 is better.

gvfsa

9 hours ago

If it depends on your use case then it’s not neck and neck…

eru

4 hours ago

Why? I'm pretty sure I can find use cases for which DeepSeek is better.

vrganj

9 hours ago

Or at a really good time, depending on who you ask.

Not everyone wants Anthropic to succeed.

jimbob45

9 hours ago

If Anthropic or OpenAI fails, the other wins the US market, no? I’m oversimplifying but…that’s a pretty realistic possibility in the face of people thinking sinking one US AI company would hurt the greater US venture into AI.

cheema33

9 hours ago

> If Anthropic or OpenAI fails, the other wins the US market, no?

I am of the opinion that they both either succeed or they both fail. They are both losing money hand over fist and their model quality is pretty close to each other. There is not a whole lot that sets them apart.

nkmnz

9 hours ago

I agree on the analysis, but not on the conclusion. What makes you think that the conditions you describe are stable grounds for a duopoly where both could be successful at the same time?

tannertech

8 hours ago

If one fails the bubble pops so you're right.

vrganj

9 hours ago

What if both fail?

tourmalinetaco

9 hours ago

If they did then Google would take over, and others would form from their competent engineers. Also, they don’t need to be profitable to succeed, they merely need to hype investors enough to ensure that their losses matter less than their growth projections.

tyrabound

5 hours ago

I wonder if Google won’t end up the last man standing in the end in the west, after Anthropic that they have a major investment and partnership with, and OpenAI crumble under their large financial burdens; while Google has a far deeper actual cash pockets, more diversified, broader and deeper bench in all related matters on top of the massive name recognition and… even if a bit brittle… a “community” of users through Gmail, Google.com, Google workspaces, etc. not to mention the TPU space and their quantum breakthrough.

vrganj

9 hours ago

What if all of them fail because the capitalist model of centralization followed by rent-seeking gets foiled by an open alternative?

That is explicitly the Chinese strategy.

Now, what if the largest capitalist economy went all-in on an AI that might no longer be sufficiently monetizable?

porridgeraisin

9 hours ago

> explicitly the Chinese strategy

Oh come on, this post sounds like one of those twitter bots. No one seriously expected them to just give away billions in R&D for free, and they arent. The chinese model companies are actively moving away from the open source setup. Even state export controls.

vrganj

7 hours ago

Let me quote the big guy himself:

> First, we should adhere to the principle of openness and win-win and boost innovation-driven development. As a new engine of world economic growth and an accelerator for the shift of growth drivers, AI is moving from the digital world into the physical world. We should seize this rare, historic opportunity to encourage open source, openness, collaboration and sharing. We should facilitate technological innovation, industrial development and scenario-based application of AI. We should make coordinated advances in the transformation and upgrade of traditional industries, the cultivation and growth of emerging industries and forward-looking planning for future industries, so that all sectors and businesses can benefit from AI

-- Keynote speech by Chinese President Xi Jinping at opening ceremony of 2026 World AI Conference

http://english.scio.gov.cn/topnews/2026-07/18/content_118605...

As a socialist myself, the above comment looks like a prime example of what Mark Fisher would call capitalist realism, the ideological belief that there is no alternative. The fact that somebody would think past quarterly earnings appears unthinkable to the committed ideologue.

They wouldn't give away R&D for free, they would give it away to destroy the capitalist mode of production, as outlined above. If that isn't a worthy goal from China's perspective, then I don't know what would be.

China isn't the US. They don't share the same ideology, strategy or conclusions.

As for the export controls, they explicitly exempt already-open models and set up an automated streamlined process for future ones.

Hamuko

9 hours ago

No? Why would either of one of those have to win? Even if you limit yourself to US companies only, Google and Meta still exist.

dgellow

4 hours ago

The others, plural, meaning Gemini, Muse, and all the open weight models hosted by a competitive landscape of US providers. There is plenty of American players in the space outside of OpenAI and Anthropic if that’s the only thing you care about.

If Anthropic and OpenAI economics are as bad as they seem to be they deserve to fail, others will take over the remaining market. It’s not a team sport

sudo_cowsay

10 hours ago

i thought open models weren't SOTA. therefore not neck and neck... in your experience, do you think it depends on your type of work?

ra

9 hours ago

They're about half a gen behind. Very impressive to get this kind of utility from a flash series model... the impact of RL seems to be behind much of the coding and agentic improvements lately.

blntechie

9 hours ago

In my experience, for coding, they are pretty much SOTA. Maybe not for other sciences and problem solving.

meerita

9 hours ago

Models are comming every 4-6 months. Anthropic, OpenAI, all of them know this.

brianjking

9 hours ago

4-6 months? Try between 8-18 days for a foundation lab release.

dudisubekti

9 hours ago

> bad time

Regardless of the performance metrics... why bad time?

Deepseek is more like a pet project with huge state backing, and they don't seem to compete with Anthropic in terms of profitability.

throwa356262

7 hours ago

Deepseek is profitable.

Anthropic is most likely not. Despite bilions of state funding and contracts.

altmanaltman

9 hours ago

It literally is not though... even DeepSeek doesn't claim that. It's just super cost-effective

baalimago

6 hours ago

I'm having trouble finding what this actually is

They're entering the search engine scene since they already have "all of the internet" stored due to the expansive LLM crawling -> training?

david-gpu

5 hours ago

It is a service for AI agents, so that they can search the web and fetch pages. It is not a product for humans.

shaunpud

5 hours ago

I "fetched" a website with hyperlinks and it only returned the text, which makes it a bit irrelevant?

fullstackwife

5 hours ago

I started to think differently about web search for agents, instead of traditional index, I would prefer to have an agentic WALKER that literally walks through pages and finds out whats needed, even if seeded with some initial web search.

jeromechoo

an hour ago

Wouldn’t searching an index be faster? Unless you’re really digging through a site for some super deeply nested link?

thenthenthen

3 hours ago

I needed to find a word on a wayback archived website. Could not. So I scraped the waybackmachine pages to .md… and found it that way… sorry Archive.org

ctippett

3 hours ago

I like the idea, but its a surefire way of bloating the context. Although I guess you'd want a subagent to do the search.

cdnsteve

3 hours ago

I use to manage Google search appliances. This feels a lot like Google old search API product. Nice to see some new players here.

apiclay

5 hours ago

This is nice. Does search include their LLM data, or does it do real-time search like Perplexity?

We are building our own in-house search API/MCP too, with anti-bot and CAPTCHA workarounds, and we plan to launch Monday morning.

gvfsa

9 hours ago

Can someone explain why fetch costs money. I thought it was your locally-running harness doing it. How can they charge you for that?

BlackRabbit1

9 hours ago

Most of your local requests will fail (with Python/Curl/whatever) if you are not using a remote-controlled Firefox/Chrome. And even then you will be bitten brutally by captchas, rendering them basically unusable for mass requests. So it's better to outsource your web search to an external service with a stable API.

The existing search machines that offer free APIs for local harnesses are super bad with low quality.

This is worth the invest as LLMs can be guided closer to existing information/code to adapt it.

hobofan

7 hours ago

> I thought it was your locally-running harness doing it.

That is rarely the case (see sibling comment for reasons). Even open source harnesses will usually guide you to connect a hosted provider for that.

Jskewel

8 hours ago

Is there any way to test this without an API key?

I am curious if my large website is in their search results. My company charges $1 per page crawled of our website, and the only traffic that worms through our firewall is using hacked residential ISPs. I really need to know who to send the (large!) bill to.

pprotas

8 hours ago

That is a really good idea, in fact I’m doing something similar. I am charging anyone reading this comment 100 bucks per letter read, please tell me who I should send the bill to.

derrida

8 hours ago

How does that work? I'm not a laywer, but I can invoice you $1 and the T.o.S of reading what I have to say are licensed under a $1 per view licence and by continued reading you consent to the terms of this license. Where can I send the bill for this comment?

I understand if you pay is a different story.

I mean ... "how does this work?"

lifeisloving

4 hours ago

Deepseeks browser platform has replaced my typical LLM search routines over the last few months.

Its extraordinary good. They arent trying to sell you things and I cab often find obscure sources/papers.

Its also great for finding research papers in chinese on a given topic.(also its free in the browser and doesn't slow my comptuer to a crawl like most webUIs by llm companies). I have it open all day nowadays.

Edit: How does this get downvoted immediately. Does Anthropic have bots just downvoting any positive sentiment about Chinese AI..

Havoc

9 hours ago

As reference point exa charges 4 bucks

andai

4 hours ago

Is AI search where the mysterious relentless crawler traffic is coming from lately?

thenthenthen

3 hours ago

Not sure, is it worse than two years ago?

larme

9 hours ago

I wonder if this is censored. If not then users in China mainland can use the api to workaround gfw (though there are much better ways)

arational

3 hours ago

Think there will be two tiers as usual, one for domestic/mainland and one for international users.

czk

2 hours ago

as expected, the typical "politically sensitive" search queries return 0 results

egress on fetch looks to be huawei cloud in hong kong

fetch tool uses a linux firefox 149 user agent with lang set to china. it runs js. could be using something like camoufox.

josteink

8 hours ago

The rate-limit given there seems really low. 5 requests per second?

I wouldn’t be able to use that API in any app with any significant number of concurrent users.

Or am I reading it wrong?

tinyhouse

5 hours ago

DeepSeek models are amazing and this search pricing is fantastic. All these companies like Parallel, Exa and others are gonna have hard time justifying their hefty valuations. Add Perplexity and many others into the mix. The demand for what they are building is there, but not the margins for the amount of money they raised.

stavros

9 hours ago

I was really excited to use this (well, I still am), but then I realized Ollama Cloud offer this too, though they don't really mention pricing. I think it's somehow rolled into their subscription.

shevy-java

9 hours ago

These prices will skyrocket eventually. AI companies need more money.

RAM prices will also probably stay very high as long as AI companis act in such a parasitic manner.

throwa356262

7 hours ago

The deepseek philosophy has always been to make a small profit or not do it at all.

I suspect they have spent months building caching infrastructure and designing intelligent fetch algorithms to get down to this number.

Or... CCP gave them a big pile of money to run this as a front for an intelligence operation. Honestly, who knows.

vitaflo

3 hours ago

I think it's important to remember that Deepseek is owned by High Flyer and it's not their core business. They don't need to make a ton of money on Deepseek to stay afloat because most of their income comes from their hedge fund business.

elendilm

4 hours ago

> The deepseek philosophy has always been to make a small profit or not do it at all.

That is a very respectable motto for any company.

0-bad-sectors

8 hours ago

Will DeepSeek pay the content creators whom pages were fetched?

nicce

7 hours ago

Is Google paying? Or why do you ask.

0-bad-sectors

7 hours ago

If you are talking about google search they direct people to the original website so the owner can monotize the traffic the way they want.

If you are talking about Gemini ( and other AI providers) no they don't they simply steel content.

I am asking to know if DS is any different because I failed to find this information on their website.

nicce

6 hours ago

Google is currently forcing that big chunk AI summary that does not redirect much traffic anymore. Afterwards, you need to navigate big chunk of ad links. Or was it first? Can't say that anymore since I have blocked them so long time ago. And then, you might be lucky to have relevant links.

timpera

4 hours ago

I really wouldn't want search engines to pay me to crawl my websites, why would you? The point of publishing stuff online is for people to access it.

fasouto

5 hours ago

Of course no, and that can disincentivize creating more content. You can copy anything and rehash it for next to nothing

ranguna

8 hours ago

Do you?

0-bad-sectors

7 hours ago

I am not a service provider. I am a site visitor that would be monetized through ads or other means